‘Why networking is way ahead to pool knowledge’
Pavel Loskot is a lecturer in multidisciplinary and nanotechnology at Swansea University. His research looks at how signal processing and networks pave the way to interdisciplinary research.
MATHEMATICAL models are an increasingly important engineering and scientific tool to forecast properties and behaviours of various systems.
In fact, we all create mathematical models of systems in our heads when trying to understand the world around us, though only a handful of us would refer to it this way.
Having spent over 15 years in modelling telecommunications systems, I changed the focus of my research to consider new areas that require truly innovative thinking, are much less saturated, and less driven by business needs and the aggressive sales and marketing rules.
I also believe traditional engineering, with its selective focus on some system aspects (electrical, mechanical, chemical and so on), has now become obsolete. A new generation of engineers needs to acquire universal problem solving skills to cope with a range of diverse problems.
One such skill is knowledge of signal processing – creating and solving the mathematical models of systems.
The real challenge is that, for every system, there are infinitely many such models possible. How simple or complex these models are is unrelated to how well they describe the observed systems; finding a simple and yet accurate model of the system can be a very difficult task.
Biological and social systems are by far more complex than any technology, and the human brain is the most complex structure known in our Universe. Until now, these systems have been studied largely by researchers with limited mathematical and engineering backgrounds.
To build models of biological and social systems, we can collect large amount of measurements; but choosing the right measurements is an art and another big challenge. Fortunately, the engineers and mathematicians can help – they can define what measurements are needed, and use sophisticated and well-established signal processing methods to create models that are just complex enough to accurately describe the observed systems.
Interestingly, many good models of complex systems including biological and social systems appear to be networks of mutually interconnected nodes. The network models are explicitly or implicitly studied in many academic disciplines.
We have semantic networks of knowledge, languages, the internet, networks of roads and utilities, atomic networks forming molecules, social networks, and others. Many seem to have similar structures as well as properties.
At Swansea University, we have established an open forum of researchers who are keen to develop interdisciplinary collaborations through shared interest in networks.
This activity, unique in the UK, could pave the way to define future academic disciplines that will consider holistic view of systems as networks of mutually interconnected subsystems.
Let me mention network models exploiting information flows. In telecommunications networks, information is represented by zeros and ones which are delivered from a sender to a destination.
However, information is much more difficult to grasp in biological and social systems. In these systems, information is no longer zeros and ones, the sender and the destination and the mechanism of information transmission are often difficult to identify and describe, information ages over time and is coupled with its spatial location (analogy of being at the right place at the right time) as well as information may have strong impact on the system (negative economic news may lead to collapse of the financial system whereas no news will ever collapse the internet network).
In summary, as an engineer with the background in signal processing, and communications engineering, I am hoping to contribute to the development of new network models to solve problems in biology, medicine and social sciences.
Contact Pavel Loskot at firstname.lastname@example.org
This article first appeared in the Western Mail‘s Health Wales supplement on 7th January 2013, as part of the Welsh Crucible series of research profiles.